Recent research has identified a form of affective illness characterized by regularly occurring fall-winter depressions which remit in the spring-summer and by atypical depressive symptoms of hypersomnia, fatigue, increased appetite, carbohydrate craving, and weight gain. This condition, known as Seasonal Affective Disorder (SAD), can often be treated effectively via exposure to bright, artificial light. Not only is response more rapid than that obtained with conventional treatments; phototherapy, unlike pharma- cotherapy, poses no risk of a lethal overdose. Despite its promise, much still is not known about the mechanisms by which phototherapy exerts its effects or the essential components of the treatment. One area of controversy concerns the spectral properties of light that are necessary and sufficient to produce an antidepressant effect. Since prolonged exposure to ultraviolet light may place a patient at risk for retinal damage, it is important to determine if a light source absent the UV wavelength would be as effective in the treatment of SAD as the full-spectrum light sources typically used. The aim of this study is to compare the efficacy of a standard light treatment, which utilizes the full light spectrum and previously has been shown to be therapeutic, with an alternative light treatment which utilizes all but the UV portion of the spectrum and has not yet been shown to be therapeutic. Forty (40) moderately depressed (Hamilton Depression score > 14), consenting outpatients who meet DSM-III-R criteria for SAD will be studied for 6-8 weeks using a randomized, double-blind, crossover design. After a 1-week extended baseline, patients will be exposed in random order to 2500 lux of Vitalite full-spectrum lights, with and without a UV filter screen, each for a period of 1 week, 2 hours/day in the early morning, with a washout period of 1 week after each treatment. Clinical response will be assessed weekly using the SIGH-SAD, HIGH-SAD, GAS, and WMI. To evaluate possible placebo effects and antidepressant effects related to partial sleep deprivation, sleep patterns will be monitored daily and assessments of expectations regarding phototherapy will be obtained prior to the start of each treatment period.